RFID interrogator family gains flexibility

Over the past few years, the telecommunications industry has seen digital software-defined radio (SDR) replace analog radio systems at a quickening pace. Moreover, programmable hardware modules are increasingly used in radio systems at different functional levels.

Elektrobit AG uses such SDR technology to facilitate implementation of the functional modules in a new family of RFID interrogators (also called readers). One of the major reasons is to use SDR as the flexible platform to implement physical-layer innovations as defined in the emerging ISO-18000-6, -5 and -4 standards. The new features in RFID can easily be implemented in SDR. Further, the architecture enables simpler control of embedded WLAN or Bluetooth interfaces for networking the interrogators.

With the growing interest in RFID for long distances (over 1 meter) it is necessary to concentrate on UHF and microwave bands working in the far field of the interrogator. Since there is also more bandwidth available, typically higher data rates and therefore faster transaction times are feasible. But longer range also implies more activated tags within the reader's electromagnetic field and interference from other interrogator units placed in the same area. Different algorithms are known to resolve multitag situations, such as binary tree search and slotted aloha principle. Listen-before-talk operation as well as rapidly changing frequency-known as frequency-agile or frequency-hopping approaches-also help to resolve multireader problems.

RFID has been taking over many functions but must work within the power limitations on the tag side. Passive tags take the energy for their chips from the RF field, which limits the reading distance to 2 to 4 kilometers (or about 1 to 2.5 miles), whereas active tags use a battery after a wake-up signal and can reach up to 10 times that distance.

While it is not a big problem to map the tag features on a tiny chip, all the new features in the interrogator can best be realized with an SDR architecture. The interrogator consists of a broadband RF front end followed by A/D and D/A conversion and a baseband processing unit (see figure). All necessary parameters, such as RF frequency and gain control, modulation and demodulation algorithms, signal generation and detection, and data encoding and decoding are realized in a programmable FPGA or DSP digital circuit. The functionality is controlled from software of the baseband processor, which at the same time is responsible for link-layer protocols (including frequency hopping/agility, listen-before-transmit operation and different anticollision algorithms) and also offers such interfaces as Ethernet, USB, Firewire or WLAN and Bluetooth.

In particular, the use of WLAN or Bluetooth interfaces must be coordinated with the interrogation process, because the interfaces operate in the same frequency band as the interrogator-for example, in the 2.4- to 2.5-GHz ISM band.

What are the benefits of using an SDR architecture instead of a standard electronic circuitry radio? It is clear that an SDR approach can cope easily with the rapidly changing world of standards, applications and new services. It is easier and faster to design a new interrogator or update existing hardware when the parameters can be adjusted just by loading a new piece of software. In Europe, for example, there is an ongoing discussion about the use of RFID in the 868-MHz band. The question is whether to change from a frequency-hopping scheme with limiting duty cycles to a listen-before-talk mode, which would mean a lot of changes in the functionality of RF and baseband parts. If under the flexible control of a processor, such a step can be made without redeveloping the hardware. Likewise, it is easier to adapt to changing modulation and channel bandwidths using digital direct synthesizers and digital filters.

Although the SDR approach is more complex than a traditional design, it offers clear advantages when tracking the fast technological changes or providing shorter time-to-market for multistandard, multifrequency solutions (for example, FCC 915 MHz or ETSI 868 MHz).

The main challenge in the design of an SDR architecture for RFID is to allocate the appropriate processing power to each function. Filtering of IF samples may require a high-speed FPGA design, while demodulation can be realized with a DSP or a fast microprocessor, such as those in the Xscale family. The media-access controller and digital-loop-carrier functionality can be implemented on standard microcontrollers.

Roland Kung is chief technical officer and Stefan Lanz is vice president, International Sales, Elektrobit AG (Bubikon, Switzerland).

See related chart